Why Psychiatry May Sometimes Need “Fuzzy” Diagnoses

“Well, while I’m here, I’ll do the work-and what’s the work? To ease the pain of living . . .”
-Allen Ginsberg

"Many words . . . don't have a strict meaning. But this is not a defect. To think it is would be like saying that the light of my reading lamp is no real light at all, because it has no sharp boundary."
-Ludwig Wittgenstein, The Blue and Brown Books

As a general proposition, most scientists and physicians prefer sharpness to fuzziness, at least when it comes to defining terms. I generally share this view, as regards psychiatric diagnosis, but only up to a point. That point is defined by the well-being of my patient-and sometimes this may call for a “fuzzy” diagnosis.

To understand why this is so, consider the following thought experiment. Imagine that we have a set of psychiatric signs and symptoms, designated as A through F. Suppose that if we select “A,B,C,D,E” as constituting our identified “Syndrome X,” we can correlate it with a specific pathophysiology and even with a specific genetic anomaly. Sounds terrific, right? And all so very “scientific”! But suppose that the syndrome, so defined, corresponds poorly to our patients’ subjective experience of distress; and furthermore, that the “ABCDE” syndrome does not lend itself well to any effective treatment.

Now imagine we start with the same A through F signs and symptoms, but we decide to group them as either “A,B,C,D,E” or as “B,C,D,E, F.” We are, in other words, making our definition of Syndrome X “fuzzier.” We now find that Syndrome X no longer corresponds as well to a particular pathophysiology or genetic defect, but that it better captures our patients’ subjective experience, and also lends itself to an effective treatment. In short, the fuzzier syndrome allows us to reduce our patients’ degree of suffering and incapacity to a greater extent than did the “sharp” ABCDE category. Which syndrome is more “real”-the “sharp” or the “fuzzy” one? Which is more pragmatically useful? And which syndrome leads us to more humane medical care, consistent with our ethical responsibilities as physicians?

Psychiatry often takes a beating for having “fuzzy” diagnoses that don’t correspond to a specific pathophysiology, laboratory test, biomarker, or genetic defect. In this regard, we are held up to considerable scorn, in contrast to the “precise” and pathophysiology-based diagnoses in the hallowed world of “real” medicine-for example, the world inhabited by pathologists, internists, and neurologists. This is certainly the conventional wisdom of the lay press, and-more frequently than I care to admit-even of many psychiatrists. But this narrative is mostly bunk.

In the first place, there is an extraordinary degree of subjectivity and “fuzziness” in general medicine, neurology, and-yes-in general pathology. I learned this recently when a family member underwent a prostate biopsy, only to be told by two pathologists that he had a “suspicious” nest of cells in one quadrant of his prostate. Suspicious? Was it cancer or not? my relative wanted to know. Nobody, including his urologist, could tell him. The specimen was sent off to an “expert” at Johns Hopkins, who, presumably, is either smarter than the first two pathologists, or nearer the Mind of God. The two puzzled pathologists were not trying to be evasive-there was simply an irreducible amount of uncertainty in their observation. Prostate cells sometimes show microscopic features that are neither clearly benign nor clearly malignant. This is because Nature itself is inherently “fuzzy,” notwithstanding our sincere desire, as diagnosticians, to “carve Nature at its joints,” as Plato put it (Phaedrus 265d-266a). Ideally, we will eventually learn what degree of cellular fuzziness we need to preserve-or cast aside-in order to provide patients like my relative with the right kind of care.

Nature may be equally uncooperative when it comes to psychiatric disorders. Our patients’ suffering may not correspond to some Platonic “essence” that has a well-defined set of necessary and sufficient criteria. Nor are such precise criteria sets necessarily more “real” than other constructions we might use as disease categories-any more than a sharply focused beam of light is more “real” than the fuzzy beam coming from Wittgenstein’s reading lamp. Of course, it is worth trying to be precise-but not necessarily “as precise as possible.” A better and more ethically defensible goal for physicians is to be as precise as is clinically useful. We are not physicists, we are physicians: our primary goal ought to be to reduce suffering and incapacity and to enhance life in all its healthy and creative dimensions. For that matter, even the physicists speak only of “probability” and “uncertainty” at the quantum level-they still can’t say precisely where that elusive electron is hiding.

To be sure: it is generally laudable, in psychiatry, to develop diagnostic criteria with high inter-rater reliability, good sensitivity and specificity, and a strong association with specific biogenetic markers. But if our diagnostic categories can’t capture our patients’ felt experience-the phenomenology of their illness-what good is “carving Nature at its joints”? If biomarkers and genotypes do not allow us to alleviate our patients’ suffering and incapacity, what service have we rendered them? I’m with the beatnik poet on this one: we are in this profession to ease the pain of living.